The use of molecular species delimitation methods is widespread in modern systematics and taxonomic research. Methods such as the Automatic Barcode Gap Discovery (ABGD, Puillandre et al., 2012), the statistical parsimony networks (e.g., Pons et al., 2006), the Bayesian Poisson Tree Processes model (bPTP, Zhang et al., 2013) and the Generalized Mixed Yule Coalescent approach (Fujisawa and Barraclough, 2013) assist in taking taxonomic decisions for challenging species, like cryptic species complexes (e.g., Fernández-Álvarez et al., 2020) or other morphologically challenging organisms. Molecular species identification methods, such DNA barcoding (Hebert et al., 2003), provide tools for identification of challenging organisms, such as undescribed ontogenetic stages of known species (e.g., Fernández-Álvarez et al., 2017, Olmos-Pérez et al., 2018a,b; Villanueva et al., 2020).

The family Sepiolidae Leach, 1817, commonly known as bobtail squids, is a very diverse group of benthic and nektonic small to medium size cephalopods. In recent years, the taxonomy and systematics of the family has been dynamic, including a few molecular phylogenetic studies (e.g., Groenenberg et al., 2009; Sanchez et al., 2019), major systematics reviews (Bello, 2019, 2020) and the description of a relatively large number of new species (e.g., de Heij and Goud, 2010; Kubodera and Okutani, 2011; Sanchez et al., 2019). In the last decade, new species have been even discovered in the Mediterranean Sea (Bello, 2013; Bello and Salman, 2015), where bobtail squid diversity has been extensively studied for more than a century (Bello, 2015, 2019). The closure of the Strait of Gibraltar is an historical process that may explain the contemporary Mediterranean sepiolid endemism and high species richness, especially in the western area (Mangold and Boletzky, 1988; Bello, 2003; Rosa et al., 2019). Sepiolid systematics and taxonomy rely mostly on the morphology of the light organs or photophores, the male copulatory organ or hectocotylus, and the female sperm storage organ or bursa copulatrix (Reid and Jereb, 2005; Bello, 2020). Three subfamilies are recognized: Heteroteuthinae Appellöf, 1898, Rossiinae Appellöf, 1898 and Sepiolinae Leach, 1817. Even though hectocotylus morphology is a reliable character, recently it was discovered that the previously recognized intraspecific variability in one species undercovered pseudocryptic biodiversity (Groenenberg et al., 2009; de Heij and Goud, 2010). Moreover, identifications based on early life stages and females are challenging and misidentifications are abundant on GenBank (Groenenberg et al., 2009), hindering identification based only on DNA barcoding.

Here, we examined most of the Mediterranean biodiversity for the family Sepiolidae, performed several molecular delimitation methods in order to assess the actual diversity of this group, and provide a solid molecular framework for future studies based on DNA identification methods. Additionally, we discovered a new species of Heteroteuthinae, which is described here using both molecular sequences and morphological characters.

All of the genera and twelve sepiolid species, covering 70% of the known specific biodiversity of the family in the Mediterranean Sea, have been barcoded and vouchered (Table 1). All the studied individuals were successfully linked with their taxonomic name and no inconsistencies arose among the newly sequenced material. Nevertheless, wrong identifications of bobtail squids are relatively frequent in GenBank, as previously reported by Groenenberg et al. (2009) and Sanchez et al. (2019). The work of Groenenberg et al. (2009) on DNA barcoding on vouchered individuals has solved some of the previously misidentified sequences. However, this work was based on animals from NE Atlantic waters and some species not studied by them remained unsolved. Here, we provided a correct barcode for many of these species, solving some of the previous problematic sequences (Table 4). Solving those problematic sequences based on morphologically identified animals is extremely important to ensure the quality of the identification based only on molecular data.

For most of the species studied here, there is a tendency of intraspecific distances below 2% and interspecific distances above 2.4–3% (Tables 2, 3). However, this pattern is not universal within the family and some exceptions occur. In fact, no clear barcode gap (Meyer and Paulay, 2005) was identified between intra- and interspecific distances, due to the presence of challenging groups, such as the clade formed by S. affinis, S. atlantica and S. intermedia. The interspecific distances of these three species are the smallest among our dataset (0.81–1.45%). In fact, the largest intraspecific p-distances among Sepiolinae (1.8% in S. robusta and 2.3% in E. scolopes, Table 3) are larger than the interspecific distances between S. affinis, S. atlantica and S. intermedia, thus existing an overlapping between intra- and interspecific distances which complicates the use of DNA barcoding methods based on genetic distances for this group of animals. The three species have very distinctive hectocotylus morphology with discrete morphologies (Figure 3): the morphological variation of those species does not overlap (Bello, 1995; Reid and Jereb, 2005; de Heij and Goud, 2010). As far as we know, no hybrids have been described among them. While S. atlantica is allopatric in reference with the other two species, S. affinis and S. intermedia both occur in Mediterranean waters (Reid and Jereb, 2005), pointing out to the presence of effective reproductive isolation mechanisms acting at least between the two Mediterranean species. For this clade, only the single threshold approach of the GMYC among all the tested molecular species delimitation methods provided the same results as the morphology (Figure 3). If molecular identifications were carried out with no further morphological information, this level of interspecific distances might be mistaken as intraspecific variation. This low level of interspecific distances might be due to recent phenomena of speciation with a fast morphological drift of key morphological characters. Recently, Costa et al. (2021) found that two morphologically different species of coastal squids that diverged in recent times were recognized as a single species by molecular species delimitation methods. Their study and ours highlights the importance of combining studies based on molecular identifications with careful morphological examinations.

Another important phenomenon hindering the direct use of bobtail squid DNA sequences for species identifications is the presence of cryptic biodiversity. Although cryptic biodiversity is an increasingly reported phenomenon in cephalopod biodiversity studies (e.g., Anderson et al., 2007, Cheng et al., 2014, Fernández-Álvarez et al., 2020), this phenomenon is comparatively unknown in bobtail squids. This is especially true for E. hyllebergi and E. tasmanica, with interspecific distances between different OTUs ranging from 3.9 to 11.6%. Remarkably, the two E. hyllebergi OTUs occur allopatrically in both Indian and Pacific coasts of the Thailand Peninsula, while E. tasmanica OTUs might occur sympatrically. It is known that the hatchling size and mode of life has an important effect on the distribution range of cephalopod species, since species with larger benthic hatchlings tend to have smaller distribution areas (Villanueva et al., 2016). Information on the way of life of bobtail squid during their first days of life is scarce (Villanueva et al., 2016: Tables 1, 2). Available information points out that Rossinae and most Sepiolinae tend to have large benthic hatchlings, although some early stages of some species reported as benthic throughout their lives can be also be found in the water column (Olmos-Pérez et al., 2018b). It is particularly remarkable that the species with larger hatchlings from the subfamily Sepiolinae, E. tasmanica (5 mm ML, Villanueva et al., 2016: Table 1), has more cryptic lineages with larger interspecific distances, accounting for four different OTUs with a divergence of 4.3–11.6% (Table 2). Interestingly, R. macrosoma has slightly larger benthic hatchlings (5.5 mm ML) and also has genetic structure, being taken as two species by several species delimitation methods. Comparisons between the sympatric geographic patterns of E. tasmanica lineages with the allopatric pattern found in Euprymna hyllebergi, a species with smaller planktonic hatchlings (2.2 mm ML), suggest that large hatchlings with direct benthic development have a more intense effect on the dispersal capacity and communication between distant cephalopod populations, and it might be one of the triggers that increase the opportunities for speciation. Rossinae are mostly exclusively benthic species, and Sepiolinae usually are usually reported as benthic species (Reid and Jereb, 2005), but they can also be found in the water column (Bello and Biagi, 1995). Absence of ontogenetic migrations have been suggested for some Sepiolinae species (e.g., Villanueva, 1995), which suggests that this restrictions to movement limit dispersal, contributing to slow down the gene flow between distant populations. This putative lower dispersal during both young and adult stages combined by the fact that some species inhabit relatively non-overlapping bathymetric ranges might have helped to trigger speciation in the Mediterranean Sea and increase their endemic bobtail species (Bello, 2019). Bobtail squids are among the cephalopods with higher tolerances to salinity changes (Mangold and Boletzky, 1988) and young stages can be found in estuarine systems (e.g., Olmos-Pérez et al., 2018b), and adults in the intertidal regions (Fernández-Álvarez, pers. obs.). Thus, it seems that salinity it is not a great limitation to bobtail squid dispersal. The Mediterranean Sea Sepiolinae fauna is characterized by the high number of endemic species (Bello, 2019), while many Atlantic species does not distribute also in the Mediterranean. It is possible that the effect of some well-known oceanographic barriers to genetic exchange, such as the Strait of Gibraltar (Pascual et al., 2017), had an important effect on the evolutionary history and current distribution of European Sepiolinae, while the differences in salinity between the Mediterranean and the Atlantic (Mangold and Boletzky, 1988) might have a comparatively smaller effect. Future population genetic studies focused on those species can answer this question.

Members of the subfamily Heteroteuthinae are exclusively nektonic and benthic-pelagic species (e.g., Orsi-Relini, 1995). Although it is known that oceanic currents can isolate populations of some pelagic squids and lead them to speciation (e.g., Fernández-Álvarez et al., 2020), the opportunities to dispersal and population connectivity are comparatively larger in oceanic environments than in shallow benthic ecosystems. Thus, large distribution areas on wide oceanic basins are commonly reported in Heteroteuthinae species (Reid and Jereb, 2005). According with Vecchione and Young (2007), there are no known morphological differences between H. dispar and H. hawaiinensis, while all the molecular species delimitation methods performed here support their treatment as different species. In the absence of known morphological differences, these two species shall be considered as members of a cryptic species complex. A third undescribed species with a not yet described morphology, Heteroteuthis sp. KER (Braid and Bolstad, 2019), form a clade with H. dispar and H. hawaiinensis. Heteroteuthis dagamensis have a large distribution ranging from the Gulf of Mexico and the South Atlantic (Judkins et al., 2016) to New Zealand (Braid and Bolstad, 2019) and North Atlantic (Taite et al., 2020). Four out six of the molecular species delimitation analyses recognized cryptic biodiversity within this species (Figure 3). The large distribution range of the species in combination with the large intraspecific divergence between New Zealand and the specimens coming from other latitudes (>2%, Table 3) suggests that some processes of speciation in its early stage might be taking place. Between those described cryptic Heteroteuthis species (H. dispar and H. hawaiinensis), undescribed new species (Heteroteuthis sp. KER) and cryptic lineages (H. dagamensis) some well-known oceanic and terrestrial barriers exists, such as the Panama Isthmus and the currents that creates the main oceanographic gyres. New combined morphological and molecular studies are necessary for solving this taxonomic problem.

It is remarkable the fact that a closer relationship exists between Sl. japonica and Se. nipponensis (2.4%) rather than between Sl. japonica and other congeneric species (12.7% with Sl. cthulhui sp. nov. and 14.3% with Sl. leucoptera). Divergence between Sepiolina petasus Kubodera and Okutani, 2011 and Se. nipponensis is 13.9%, similar to that reported between species of the genera Sepiolina and Heteroteuthis. These data suggest that the current generic assignations to genera in Heteroteuthinae species are not fully molecularly supported and should be revised in future combined morphological and molecular studies, as already suggested by Allcock et al. (2014).

All molecular species delimitation analyses consistently identified Sl. leucoptera and Sl. cthulhui sp. nov. as different species (Figure 3), also the COI divergence of 3.5% is typical for different species in bobtail squids. This difference is consistent with interspecific levels in many other invertebrates, such as nemerteans (Fernández-Álvarez and Machordom, 2013), land planarians (Lago-Barcia et al., 2015), crustaceans (Robles et al., 2007), and other cephalopods (Gebhardt and Knebelsberger, 2015; Fernández-Álvarez et al., 2020). Besides, the morphological comparisons between Sl. leucoptera and Sl. cthulhui sp. nov. showed a few differences in key morphological characters that until now remained overlooked, such as the length of the tentacle, the width of the head and slight differences in the sexual modifications of arm II in mature males. In general, it seems that the modifications of the arm II in Sl. cthulhui sp. nov. are less pronounced than in Sl. leucoptera. Even so, the differences between both species are few, likely representing a shallow morphological drift and a recent speciation phenomenon. Species involved in those first steps are commonly referred as in the “gray speciation zone” (Roux et al., 2016). As we showed with S. affinis, S. atlantica and S. intermedia, as well as the molecular data provided for the descriptions of Sl. japonica and Sp. nipponensis, these situations are likely to be frequent in bobtail squids. Based on this observation, the specific status of distant populations with mild levels of genetic divergence, which are typically identified as intraspecific levels of divergence, should be taken with care.

We took special care of including as many morphological characters as possible to avoid overlooking possible morphological differences between the congeneric Sl. cthulhui sp. nov. and Sl. leucoptera. That is the case of beak and radulae morphologies, which are rarely used in bobtail squid taxonomy (e.g., Kubodera and Okutani, 2011; Sanchez et al., 2019), and spermatophores. The use of spermatophores in cephalopod taxonomy should be taken with caution, as our study also shows. It is known that spermatophore size depends on the size of the male that produces it and since they can accumulate spermatophores for long periods of time, the same individual can hold spermatophores of a huge range of sizes and morphologies, according with the somatic size of the squid when each spermatophore was formed (Hoving et al., 2010; Cuccu et al., 2014). The number of spermatophores stored by the male would vary according with when it mated the last time (if anything at all) and so will do the size of those structures. In the present study we found 353–358 spermatophores in the spermatophoric sacs of Sl. leucoptera, while 236 (ICMC000165) and 11,645 (ICMC000166) spermatophores were found in the two examined specimens of Sl. cthulhui sp. nov. It is interesting to point the fact that the specimen ICMC000165, with a lower number of spermatophores, was also the same individual with a narrowest range of spermatophore size. Particularly, the specimen ICMC000166 not only showed the largest range of spermatophore size: it also showed empty and aberrant spermatophores. It is known that in the beginning of their reproductive life, cephalopod males produce tentative spermatophores (Nigmatullin et al., 2003) and in this case both functional and tentative spermatophores were present in the spermatophoric sac. It is not known to us if this was related with the absence of mating events by this individual for any circumstance or if it is just an aberrant individual. Pelagic cephalopods might have difficulties to find conspecifics to mate (Hoving et al., 2012), or just the opposite situation (Fernández-Álvarez et al., 2018; Hoving et al., 2019), so species with promiscuous mating systems are expected to hold a less variable morphology in the spermatophores stored in the spermatophoric sac. Special attention should be taken to the characteristics of a cephalopod species sex life if spermatophores are going to be used as a taxonomic character.

In this study we provided new sequence data for most of the Mediterranean bobtail squid species and added other relevant sequence data from species from other parts of the world. The combined use of several molecular species delimitation methods consistently showed a picture of hidden biodiversity. On the one hand, most of those methods failed to accurately assess the actual biodiversity of some morphologically different species, hindering the use of molecular data for species identification in the absence of morphological data. On the other hand, cryptic biodiversity was detected among members of the same nominal species, pointing to the fact that in some cases, speciation phenomena might be occurring without a parallel morphological evolution. It is also possible that the comparatively low number of taxonomists working on bobtail squids has hindered the discovery of morphological differences between them or that some morphological differences had been overlooked, as it happened with the new Stoloteuthis species described in this work. The Mediterranean Sea is one of the more diverse and the better studied areas for members of the family Sepiolidae, with a literature production spanning through more than a century (Bello, 2015). Despite this intense biodiversity and taxonomic research, two new Mediterranean species were described in the last few years (Bello, 2013; Bello and Salman, 2015). Here, a new species of Mediterranean Stoloteuthis, previously misidentified, was discovered and described, both using molecular and morphological methods. It is also remarkably that Olmos-Pérez et al. (2018b) found a species (“Heteroteuthinidae” sp.) in European waters whose sequences cannot be assigned to any sequenced species of bobtail squid, while Sanchez et al. (2019) discovered two new species of the genus Euprymna in Pacific waters. All these recent new species discoveries stress the need of new taxonomic studies in both benthic and pelagic bobtail squids on a worldwide basis.

The datasets presented in this study can be found in online repositories. The names of the repository/repositories and accession number(s) can be found at: NCBI GenBank (accession: MW260131-MW260134).

Ethical review and approval was not required for the animal study because the cephalopods we worked with for this study were all dead before research began.

FÁF-Á, PS, and RV conceived the study, participate in sample collection and identification, and performed morphological examinations. FÁF-Á performed the phylogenetic analyses, data curation, and prepared the first draft. All authors contributed to the manuscript preparation and revision and read and approved the submitted version.

The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.